In previous years we had disrupted the genes encoding the pertussis toxin sensitive G proteins Gi2, Gi1, Gi3 and Go. Double knockouts involving Gi2 and Go are lethal. During this past year we have added two conditional knockout lines by floxing Gi2 and Go. We expect to learn from combining Gi2 with Gi3 and enhanced survival of Go KO mice by removing the floxed genes at various times after birth. Breeding programs have been set up to add cre recombinase under several specific promoters so as to remove the genes both generally in all issues or in specific cell types such as in dopaminergic neurons to remove Go or lymphocytes to remove Gi from Gi3 KO mice. [unreadable] [unreadable] We have obtained founder mice carrying the Gs-alpha encoding Gnas complex locus (120kb) with an in-frame fluorescent protein (red and green) fused to exon 1 of Gs. Surprisingly the first three founders (carrying Gnas-GFP) did not breed and have low sperm count. We have two founders for the Gnas-RFP transgene and are curious to see if the phenotype is the sam. We will attempt artificial insemination to keep the trait alive and elucidate the phenotype and also study imprinting. An allele specific PCR has been set up to study Gs-alpha imprinting in tissues.[unreadable] [unreadable] Good progress has been made this year to set up a method that allows us to survey unmethylated CpGs at a genome-wide scale. The approach was eventually changed from the one sketched out in last year's annual report by switching from MmeI to EcoP15I as the tag-tag-removing enzyme, and instead of creating di-tags followed by concatenation, we amplify the tags directly by PCR and sequence the library so-made using the Solexa technology. IN a first run Illumina returned 39 million 33nt sequences that were derived from a potential 2-4 million unmethylated CpGs. We are improving the library construction conditions and learning how to analyze the data. While NIEHS's Biostatistics Branch is well prepared to provide statistical support for toxicology studies of the type performed for the National Toxicology Program, it is poorly and ill-equipped to provide the informatics support to DIR's PIs that need to analyze results obtained with new generation high throughput sequencing machines, to which Solexa belongs. We are now, finally, in position to test the hypothesis whether the pattern of unmethylated CpGs can established for blood bone cells and whether these patterns can be used to report on environmental exposures leaving a footprint in the stem cell niche of the hematopoietic compartment of the bone marrow. Additional information to be gathered will contribute to the description of differentially methylated domains or regions (DMRs) and whether methylation status can inform about allelic exclusion phenomena of which imprinting is but one of many forms in which allelic exclusion is manifested. [unreadable] [unreadable] Other ongoing studies focus on properties of Gs-alpha mutants that may inform on the molecular mechanism by which receptors activate Gs. The first mutant analyzed was Gs-alpha R265E which showed commonalities to what was seen with cognate mutations in transducing alpha, but also features that are unique to Gs. A second mutant Gs-alpha T204A revealed that the conformational change that confers the ability of Gs-alpha to activate adenylyl cyclase (effector activating function) is independent of the conformational change responsible for activation of its GTPase activity (auto-turnoff function. We are crystallizing the cognate mutation in a transducin alpha subunit to better understand the atomic basis of the changed properties. [unreadable] [unreadable] Finally, during this last year we have established a small crystallography group of three persons with the Transmembrane Signalling Group: staff scientist Dr. Yanshun Liu and visting IRTA fellows Ankita Roy and Yinghao Zhang. The primary goal of this group will be to co-crystallize an engineered form of the trimeric G protein transducin with an engineered form of the G protein-coupled receptor rhodopsin. [unreadable] [unreadable] We continue collaborating with extramural scientists in the analysis of the phenotypes that arise in G protein deficient mice. The most notable finding has been that Go2 KO mice have a diminished tendency to become addicted to cocaine and that this effect is due to altered dopmine storage in dopminergic neurons.